Research On Two-product Newsboy Problem:Based On Prospect Theory And Information Updating

Considering the decision-maker’s actual decision is always deviated from the theoretical optimal solution in risk-neutral, this paper applied the prospect theory and information updating method to analyze the order model of two products which are perishable product under random demand. Furthermore, we combined modeling with numerical example to study the risk decision problem. Specifically, the main contents of this paper can be summarized as the following three aspects.First, based on the framework of prospect theory, we analyzed the value function, subjective probability and decision weighting function while considering two products under random demand, and then the order model is established. Compared with other single product research, this model considers two products and lays the foundation of multi-production research. It makes an important progress in the research of multi-product newsboy problem while using prospect theory. The order problem of two products based on prospect theory is analyzed from many different aspects in the numerical example (e.g., salvage value, psychological reference points, risk-attitude coefficient). The results show that this model describes the retailers’actual order behavior much better.Second, we introduced information updating into the previous two-product ordering model. The retailer has information updating ability in the lead time while more later he order, more demand information he can use to improve his prediction accuracy, but the extra cost is higher. So we established order model to show when and how much should the retailer decide to order to the two products while he is loss aversion. Then the numerical example indicates that the change trend of two products’ optimal order quantities along with the psychological reference point is the same with the circumstance without imformation updating. But the degree of change is significantly weakened by imformation updating. Optimal order time doesn’t have obvious change along with the psychological reference point. Howeve, as loss aversion coefficient increases, the optimal order time will delay. Study also shows that the information of high-risk product’s demand is more valuable to retailer compared with the infonnation of low-risk product.Third, we extended previous problem to a two-stage order problem that the retailer is loss aversion and he sales two products while he can updates the information of demand. And we used the bayesian infonnation updating model based on Brownian movement to update the demand. Then we established the order model of two products that saled in two stages while the retailer is loss aversion. Furthermore, we applied Monte Carlo method to solve the four dimensions of the integral problem, and the solve algorithm is designed.